Agentic AI earns trust when autonomy has boundaries. Define tool permissions, approval points, and an audit trail before connecting agents to real work. #AI #AIAgents #ResponsibleAI
November 1, 2026 at 2:30 PM
Agentic AI earns trust when autonomy has boundaries. Define tool permissions, approval points, and an audit trail before connecting agents to real work. #AI #AIAgents #ResponsibleAI
The best AI workflow isn’t fully automated. Let AI handle repeatable steps, then give people clear context for exceptions and consequential decisions. #HumanAI #ResponsibleAI #AI
November 1, 2026 at 4:01 PM
The best AI workflow isn’t fully automated. Let AI handle repeatable steps, then give people clear context for exceptions and consequential decisions. #HumanAI #ResponsibleAI #AI
When AI supports decisions about people, affected people need understandable explanations and a way to seek review. Good systems make accountability part of the design. #AI #HumanCenteredAI #ResponsibleAI
November 1, 2026 at 8:03 PM
When AI supports decisions about people, affected people need understandable explanations and a way to seek review. Good systems make accountability part of the design. #AI #HumanCenteredAI #ResponsibleAI
AI agents are moving from demos into real workflows. The next challenge is making their actions traceable, bounded, and easy for people to review. #AIAgents #ResponsibleAI
November 1, 2026 at 5:53 AM
AI agents are moving from demos into real workflows. The next challenge is making their actions traceable, bounded, and easy for people to review. #AIAgents #ResponsibleAI
An AI pilot should prove more than a polished demo. Track accuracy, time saved, escalation rates and user feedback in the real workflow before expanding access. #AI #ResponsibleAI #EnterpriseAI
November 2, 2026 at 5:51 PM
An AI pilot should prove more than a polished demo. Track accuracy, time saved, escalation rates and user feedback in the real workflow before expanding access. #AI #ResponsibleAI #EnterpriseAI
AI agents can speed up work, but trust comes from clear boundaries: what they may do, when a person must approve, and how every decision is recorded. #AI #AIagents #ResponsibleAI
November 1, 2026 at 8:02 PM
AI agents can speed up work, but trust comes from clear boundaries: what they may do, when a person must approve, and how every decision is recorded. #AI #AIagents #ResponsibleAI
To get a reasonable confidence level in the population I need at least 100 responses. Please can you help #DataGovernance #AIGovernance #DataManagement #ResponsibleAI #Leadership #DigitalTransformation #MicrosoftFabric
Can you spare 2 minutes? I'm conducting research into data and AI governance readiness. Anonymous, 10 yes/no questions, and every response helps forms.cloud.microsoft/e/bdh3fMN7VP #DataGovernance #AIGovernance #DataManagement #ResponsibleAI #Leadership #DigitalTransformation
Microsoft Forms
forms.cloud.microsoft
November 1, 2026 at 9:27 PM
To get a reasonable confidence level in the population I need at least 100 responses. Please can you help #DataGovernance #AIGovernance #DataManagement #ResponsibleAI #Leadership #DigitalTransformation #MicrosoftFabric
Good AI governance is practical: assign an owner, document important decisions, and review performance as the system changes. #AI #AIGovernance #ResponsibleAI
November 3, 2026 at 6:30 AM
Good AI governance is practical: assign an owner, document important decisions, and review performance as the system changes. #AI #AIGovernance #ResponsibleAI
AI governance works best when it is part of everyday work: clear ownership, documented risks, human review where needed, and a way to challenge outcomes. #AI #AIGovernance #ResponsibleAI
November 1, 2026 at 8:03 PM
AI governance works best when it is part of everyday work: clear ownership, documented risks, human review where needed, and a way to challenge outcomes. #AI #AIGovernance #ResponsibleAI
Can you spare 2 minutes? I'm conducting research into data and AI governance readiness. Anonymous, 10 yes/no questions, and every response helps forms.cloud.microsoft/e/bdh3fMN7VP #DataGovernance #AIGovernance #DataManagement #ResponsibleAI #Leadership #DigitalTransformation
Microsoft Forms
forms.cloud.microsoft
November 1, 2026 at 9:51 AM
Can you spare 2 minutes? I'm conducting research into data and AI governance readiness. Anonymous, 10 yes/no questions, and every response helps forms.cloud.microsoft/e/bdh3fMN7VP #DataGovernance #AIGovernance #DataManagement #ResponsibleAI #Leadership #DigitalTransformation
Good AI governance is practical: name an accountable owner, document important decisions, and monitor performance as models and data change. #AIGovernance #AI #ResponsibleAI
November 3, 2026 at 7:19 AM
Good AI governance is practical: name an accountable owner, document important decisions, and monitor performance as models and data change. #AIGovernance #AI #ResponsibleAI
A polished demo is not an AI evaluation. Test the system on real tasks, unusual inputs, failure cases, security risks, and who is affected when it gets something wrong. #AI #AItesting #ResponsibleAI
November 2, 2026 at 7:47 PM
A polished demo is not an AI evaluation. Test the system on real tasks, unusual inputs, failure cases, security risks, and who is affected when it gets something wrong. #AI #AItesting #ResponsibleAI
Trustworthy AI should make limits visible. When a system is unsure, people need to know what it can support and what needs verification. #AI #DigitalTrust #ResponsibleAI
November 3, 2026 at 6:31 AM
Trustworthy AI should make limits visible. When a system is unsure, people need to know what it can support and what needs verification. #AI #DigitalTrust #ResponsibleAI
Reliable AI depends on more than model quality. Teams need representative evaluations, a way to catch failures after launch, and clear ownership when something goes wrong. #ResponsibleAI #AIEvaluation
October 31, 2026 at 6:10 AM
Reliable AI depends on more than model quality. Teams need representative evaluations, a way to catch failures after launch, and clear ownership when something goes wrong. #ResponsibleAI #AIEvaluation
Responsible AI is built through everyday decisions: representative testing data, documented limits, privacy safeguards, and a way to report problems. Good practice is operational. #AI #ResponsibleAI #AIGovernance
November 2, 2026 at 8:31 PM
Responsible AI is built through everyday decisions: representative testing data, documented limits, privacy safeguards, and a way to report problems. Good practice is operational. #AI #ResponsibleAI #AIGovernance
AI agents work best with a defined role, limited permissions, and a clear handoff to people when uncertainty rises. Boundaries make autonomy useful. #AIAgents #ResponsibleAI #Governance
November 3, 2026 at 7:19 AM
AI agents work best with a defined role, limited permissions, and a clear handoff to people when uncertainty rises. Boundaries make autonomy useful. #AIAgents #ResponsibleAI #Governance
AI is most valuable when it helps people make better decisions—not just faster ones. Define the decision, surface the evidence, and keep accountability clear. #AI #ResponsibleAI #DecisionIntelligence
November 3, 2026 at 6:27 AM
AI is most valuable when it helps people make better decisions—not just faster ones. Define the decision, surface the evidence, and keep accountability clear. #AI #ResponsibleAI #DecisionIntelligence
AI initiatives move faster when teams start with one workflow, define what good looks like, and keep a human accountable for exceptions. Measure the value before scaling. #AI #EnterpriseAI #ResponsibleAI
November 1, 2026 at 5:26 AM
AI initiatives move faster when teams start with one workflow, define what good looks like, and keep a human accountable for exceptions. Measure the value before scaling. #AI #EnterpriseAI #ResponsibleAI
AI governance works best when it is part of delivery: named owners, clear approval points, and evidence attached to decisions. That gives teams room to move while keeping accountability visible. #AIGovernance #ResponsibleAI
November 1, 2026 at 11:50 AM
AI governance works best when it is part of delivery: named owners, clear approval points, and evidence attached to decisions. That gives teams room to move while keeping accountability visible. #AIGovernance #ResponsibleAI
Human oversight should match the risk. Routine tasks can move quickly; high impact decisions need stronger review, traceability, and a clear route to challenge outcomes. #AI #ResponsibleAI #Governance
November 3, 2026 at 9:17 AM
Human oversight should match the risk. Routine tasks can move quickly; high impact decisions need stronger review, traceability, and a clear route to challenge outcomes. #AI #ResponsibleAI #Governance
Looking forward to this important discussion this afternoon
⚗️when digital tools might serve local priorities, and when they might not, and knowing how to know
#AIforPeace #Peacebuilding #StockholmForum #ResponsibleAI
👇
⚗️when digital tools might serve local priorities, and when they might not, and knowing how to know
#AIforPeace #Peacebuilding #StockholmForum #ResponsibleAI
👇
If you are at the Stockholm Forum on Peace and Development today, join us this afternoon for a practical workshop on “Surfing the AI wave: How can organisations integrate PeaceTech?"
📅 3 November, 2025
⏰16:00 - 17:30
📍Arkaden
📅 3 November, 2025
⏰16:00 - 17:30
📍Arkaden
November 3, 2026 at 9:32 AM
Looking forward to this important discussion this afternoon
⚗️when digital tools might serve local priorities, and when they might not, and knowing how to know
#AIforPeace #Peacebuilding #StockholmForum #ResponsibleAI
👇
⚗️when digital tools might serve local priorities, and when they might not, and knowing how to know
#AIforPeace #Peacebuilding #StockholmForum #ResponsibleAI
👇
🧪 When AI breaks a promise, can organisations blame the algorithm and escape accountability?
My new paper, “The Algorithmic Scapegoat,” examines how AI autonomy can become a convenient excuse for organisational responsibility evasion.
doi.org/10.1016/j.jr...
#ResponsibleAI
My new paper, “The Algorithmic Scapegoat,” examines how AI autonomy can become a convenient excuse for organisational responsibility evasion.
doi.org/10.1016/j.jr...
#ResponsibleAI
Redirecting
doi.org
October 29, 2026 at 9:38 PM
🧪 When AI breaks a promise, can organisations blame the algorithm and escape accountability?
My new paper, “The Algorithmic Scapegoat,” examines how AI autonomy can become a convenient excuse for organisational responsibility evasion.
doi.org/10.1016/j.jr...
#ResponsibleAI
My new paper, “The Algorithmic Scapegoat,” examines how AI autonomy can become a convenient excuse for organisational responsibility evasion.
doi.org/10.1016/j.jr...
#ResponsibleAI
A strong AI evaluation suite tests more than accuracy. Include edge cases, changing data, user behavior, failure recovery, and escalation to a human. Reliability is a product feature. #AI #ResponsibleAI
October 31, 2026 at 9:00 PM
A strong AI evaluation suite tests more than accuracy. Include edge cases, changing data, user behavior, failure recovery, and escalation to a human. Reliability is a product feature. #AI #ResponsibleAI